© 1997 American Heart Association, Inc.
Articles |
Cross-Trait Familial Resemblance for Body Fat and Blood Lipids: Familial Correlations in the Quebec Family Study
From the Division of Kinesiology, Laval University School of Medicine, Ste-Foy, Québec (L.P., J.P.D., C.B.); Division of Biostatistics (T.R., D.C.R.) and Department of Psychiatry and Genetics (D.C.R.), Washington University School of Medicine, St Louis, Mo; Lipid Research Center, Laval University Medical Research Center (J.P.D.).
Correspondence to Louis Pérusse, PhD, Division of Kinesiology, Physical Activity Sciences Laboratory, Laval University–PEPS, Ste-Foy, Quebec, G1K 7P4 Canada. E-mail louis.perusse{at}kin.msp.ulaval.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Abstract In an attempt to better understand the genetic basis of the metabolic syndrome, we have undertaken a series of studies on the familial aggregation in the clustering of the coronary heart disease risk factors which characterize this syndrome. In the present study, the hypothesis of shared genetic (pleiotropy) and/or environmental factors between body fat and blood lipids is investigated by examining cross-trait (eg, father's body fat with his son's blood lipid) familial resemblance between 4 indicators of body fat (body mass index [BMI], sum of 6 skin folds [SF6]) and fat distribution (the ratio of the trunk to extremity skin folds adjusted [TER-sf] and unadjusted [TER] for SF6), and 5 blood lipid variables (total cholesterol [CH], triglycerides [TG], cholesterol associated with high-density lipoproteins [HDL], the CH/HDL ratio and the difference between CH and HDL [CH-HDL]) measured in 1239 individuals from 309 families participating in the Quebec Family Study. A bivariate correlation model was used to obtain maximum likelihood estimates of cross-trait spouse, parent-offspring, and sibling correlations after adjustment of body fat and lipid data for the effects of age, separately in the four sex-by-generation groups. Likelihood ratio tests revealed the presence of significant (P<.05) cross-trait resemblance between body fat (BMI and SF6) and all lipid traits except CH and also between fat distribution (TER and TER-sf) and CH/HDL and CH-HDL. Only sibling cross-trait correlations were significant for all body fat-lipid pairs of measures, with bivariate familiality estimates (ie, shared genetic and/or environmental factors) ranging from 8% to 40%. Although the hypothesis of genetic pleiotropy cannot be ruled out from the pattern of cross-trait correlations found in the present study, we conclude that environmental factors shared within sibships are probably more important than common genes in determining the covariation between body fat and blood lipids.
Key Words: body fat • blood lipids • pleiotropy
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Numerous epidemiologic studies performed over the last decade have shown that adults of both sexes frequently exhibit a clustering of CHD risk factors that includes hypertension, dyslipoproteinemias, upper-body obesity, and diabetes.1 This clustering of metabolic abnormalities, which appears to be characterized by hyperinsulinemia resulting from insulin resistance, has been variously referred to as syndrome X,2 the deadly quartet,3 the insulin resistance syndrome,4 the metabolic syndrome,5 or the plurimetabolic syndrome.6 This cluster could explain in part the increased risk of atherosclerosis and CHD frequently reported among individuals with excessive accumulation of adipose tissue in the truncal-abdominal area.7–11 Several of the metabolic abnormalities were found to be important correlates of obesity, especially abdominal or upper body obesity.4,12-14 It is therefore suggested that amount of body fat as well as upper-body or abdominal fat are important features of the metabolic syndrome.
Results from genetic epidemiology studies suggest that the various phenotypes associated with causes and manifestations of the metabolic syndrome are influenced by genetic factors.15 For example, we have shown that body fat and regional fat distribution16 as well as blood lipids and lipoproteins17 were significantly influenced by genetic factors in the Quebec Family Study. Few attempts have been made to determine whether shared genetic and/or environmental factors could be responsible for the clustering of metabolic abnormalities encountered in this syndrome. Based on twin data, Carmelli et al18 reported concordance rates of 31.6% in MZ twins compared to 6.3% in DZ twins for the familial clustering of obesity, hypertension, and diabetes as assessed by questionnaire. Using a multivariate path analysis model fitting of the data in which the clustering of obesity, diabetes, and hypertension was assumed to be mediated by a latent factor, they showed that 59% of the variance in this latent factor was accounted for by genetic factors.18 More recently, the genetic and environmental etiologies of 5 traits associated with the metabolic syndrome were investigated in a sample of 289 elderly (52 to 86 years) MZ and DZ twins.19 The phenotypes investigated included BMI, serum levels of triglycerides and HDL-cholesterol, systolic blood pressure, and an indicator of insulin resistance derived from fasting levels of glucose and insulin. The cross-twin correlations were higher in MZ twins compared with DZ twins, suggesting a shared genetic basis in the covariation between all components of the syndrome. Furthermore, the authors found evidence for a single latent genetic factor common to the 5 phenotypes investigated.19 Results from these two twin studies suggest that shared genetic factors play a role in the clustering of the morbidities associated with the metabolic syndrome.
Familial correlations (spouse, parent-offspring, and sibling) could also be used to investigate the genetic basis of the metabolic syndrome. In the traditional univariate case, the pattern of familial correlations suggests whether the trait under investigation is heritable. In the bivariate case, the pattern of cross-trait familial correlations, eg, trait 1 in a parent with trait 2 in an offspring, provides an indication about the contribution of shared genes and/or environmental factors for the two traits. For example, significant parent-offspring and sibling cross-trait correlations in the presence of a nonsignificant spouse cross-trait correlation suggest a common genetic basis for the two traits. Significant spouse cross-trait correlations in addition to parent-offspring and sibling cross-trait correlations suggest that shared environmental factors could also be involved. We have undertaken a series of investigations aimed at determining the role of genetic and environmental factors in the covariation observed among some of the phenotypes involved in the metabolic syndrome by computing bivariate familial correlations. We have already shown a common familial basis in the covariation between body fat and blood pressure20 and between body fat and fasting plasma insulin.21 The present study is aimed at determining the pattern of familial resemblance between blood lipid and lipoprotein phenotypes and body fat and subcutaneous fat distribution.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sample
The Quebec Family Study (QFS) consists of a random sample of 375 families of French descent living within 80 km around Québec City. These families were recruited through the media during the years 1978 to 1981 (Phase I) to study the genetic effects on several physiological and biochemical traits. The total sample in Phase I includes 1630 individuals comprising 727 parents ranging in age from 30 to 59 years of age and 903 offspring, 8 to 26 years old. None were diabetic or treated for cardiovascular disease. The average socioeconomic status of these families based on the occupation of the father was comparable to the general French-Canadian population.22 In addition to traditional nuclear family members (ie, parents and their singleton offspring), there were twins, adoptees, cousins, stepparents, etc. The complete sample was used for the purposes of data adjustments (ie, age and sex corrections as described below), while only traditional nuclear families were used in the correlation analyses. The sub-sample used in the familial correlation analyses included 1239 individuals from 309 nuclear families with sibship size ranging from 1 to 4.
Body Fat and Blood Lipids Measurements
A variety of physiological and behavioral
measurements was obtained during a 1-day visit of the families to the
laboratory. Measures relating to body fat included weight, height, and
skinfold thicknesses. Body weight and height were measured without
shoes in light clothing. Six measures of skinfold thicknesses
(suprailiac, subscapular, abdominal, medial calf, biceps, triceps) were
obtained on the left side of the body with a Harpenden skinfold caliper
following the procedures recommended by the International Biological
Program.23
Four indices of body mass, body fat, and fat distribution were computed from these body fat measures. The body mass index (BMI) was computed as weight (kg)/height (m2). Two variables were extracted from the 6 skin folds; the sum of all 6 (SF6) and the trunk to extremity skinfold ratio [TER = (suprailiac + subscapular + abdominal)/(medial calf + bicep + tricep)] were used as indicators of subcutaneous fat and preferential deposition of subcutaneous fat on the trunk rather than extremities. Moreover, the TER was adjusted for total subcutaneous fat (SF6) using regression analysis (TER-sf) to assess preferential deposition of body fat on the upper body independently of the total amount of subcutaneous fat. The regression analysis consisted in a stepwise procedure, extracting the standardized residuals from the regression of TER on up to a cubic polynomial in SF6.
Although the correlations between BMI and SF6 are moderately high (0.6 to 0.8), both indices were included in the analyses because BMI is a widely used indicator of obesity. High correlations were also observed between TER and TER-sf (about 0.9). Both variables were also included since TER is an indicator of the overall pattern of body fat distribution, while TER-sf takes into account the total amount of subcutaneous fat.
Serum blood lipid levels were determined from blood samples collected early (about 8:00 AM) in the morning after a 12-hour overnight fast. Details regarding blood drawing and blood lipid determinations may be found elsewhere.24 CH, cholesterol associated with HDL, and TG were determined enzymatically with commercial kits as described in detail elsewhere.17,24 Two other variables were derived from these blood lipid measurements: the CH/HDL ratio and the difference between CH and HDL (CH-HDL) used as indices of non–HDL- and apoB-associated cholesterol, respectively.
Table 1
gives the means and standard
deviations (SD) of the unadjusted variables, separately in the four
sex-by-generation groups (fathers, mothers, sons, and daughters).
Generation and sex differences are observed for most variables. In
general, there are higher values in parents than in offspring, except
for HDL values in males, which are higher in offspring than in
parents.
|
Data Adjustments
Each of the four body fat and five lipid variables were
adjusted for the effects of age in both the mean and variance, using a
stepwise multiple regression procedure. Given the significant group
differences in the means, these adjustments were conducted separately
by sex and generation groups. First, a given measure was regressed on
up to a cubic polynomial in age in a stepwise manner, retaining only
those terms which were significant at the 5% level (mean regression).
The residual from this mean regression was retained. Second, age
effects in the variance (heteroscedasticity) were examined by
regressing the square of the residual obtained above (or the log of the
squared residual) on another polynomial in age in a stepwise manner and
retaining terms significant at the 5% level (variance regression). The
predicted score from the variance regression was retained. The final
phenotype used in the correlation analysis was computed
for all individuals by using the best regression models. More
specifically, the final phenotype was the residual from the
mean age regression divided by the square root of the predicted score
from the variance regression, and standardized to ensure zero mean and
unit variance. Also, if there were no age effects, then the final
phenotype used in the correlation analysis was simply
the standardized score (zero mean and unit variance). Since all of
these data adjustments were conducted separately within each of the
fathers, mothers, sons, and daughters, the means and variances for the
final phenotypes were equal for all groups.
Age regression results may be found elsewhere for BMI,25 for SF626 and for TER-sf.20 Data adjustment of the lipids has not been previously reported. For CH, age effects in the mean were significant for fathers (linear age accounting for 3.2% of the variation), mothers (age3 accounting for 6.3%), sons (age, age2 accounting for 6.2%) and daughters (age, age3 accounting for 6.5%). Heteroscedasticity was noted only in daughters (age accounting for 1.4%). For TG in fathers, neither mean nor variance age effects were found; in mothers, sons, and daughters a linear term in age accounted for 5.7%, 5.2%, and 2.6% of the variation, respectively; heteroscedasticity was noted only in mothers (linear term accounting for 2.8%). Finally, for HDL, no age effects in either the mean or variance was found for fathers, mothers, or daughters; in sons, age and age3 terms accounted for 21.9% of the mean variation and heteroscedasticity was also noted (linear term accounting for 1%).
Bivariate Family Correlation Model
Details of the bivariate familial correlation model used in this
study may be found elsewhere.20 In summary, the
bivariate model is a simple extension of the univariate
familial correlation model involving 4 types of individuals (F
indicates fathers, M, mothers; S, sons; D, daughters) and leading to 8
intraindividual correlations: 1 spouse (FM), 4 parent-offspring (FS,
FD, MS, MD), and 3 sibling (SS, DD, SD). In the bivariate model, each
of these 8 correlations becomes a matrix of correlations. The structure
of each matrix is that within-trait comparisons are on the diagonals,
while cross-trait correlations are on the off-diagonals (see Table 2 for details).
|
In the bivariate model, the cross-trait correlations are the ones we are interested in. Element notation is used in presenting the correlations. For example, the term f1m2 denotes a spouse cross-trait correlation (father's body fat and mother's lipid), the cross-trait parent-offspring correlation f1s2 represents father's body fat and son's lipid, and f12 is the cross-trait intraindividual correlation within fathers.
The computer program SEGPATH27 was used to estimate the familial correlations by maximum likelihood methods. More details concerning the model and computer program are given elsewhere.20
Hypothesis Testing
A general model (all 34 correlations shown in Table 2
is
estimated for each bivariate pair of variables. Several hypotheses
can be tested against this general model, but only tests on sex and
generation differences in the correlations as well as on the
significance of the cross-trait correlations were considered. The
parameter reductions involved in each of the reduced models
tested are given in the Appendix. Sex and generation differences were
tested in models 2, 3, and 4. In model 2, the hypothesis of no sex
differences in offspring was tested by equating the correlations (7
sibling, 8 parent-offspring, and 1 intraindividual) involving sons and
daughters. In model 3, no sex differences in either parents or
offspring were allowed, leading to a reduction of 7 sibling, 12
parent-offspring, 1 spouse, and 2 intraindividual correlations. In
model 4, the hypothesis of no sex nor generation differences is tested
leading to a reduction of 23 sibling and parent-offspring correlations,
1 spouse correlation, and 3 intraindividual correlations. Tests on
cross-trait correlations are listed in models 5 to 10. Models 5 and 6
test the hypotheses of no cross-trait sibling and parent-offspring
correlations, respectively, by fixing the cross-trait correlations of
the corresponding matrices to zero (see Table 2 and appendix). In model
7, the hypothesis of no cross-trait correlation in either the siblings
or parent-offspring is tested, while, in model 8, the two cross-trait
spouse correlations are fixed at zero. In model 9, the cross-trait
correlations of the four intraindividual matrices (see Table 2) are
fixed at zero to test the hypothesis that there is no correlation
between body fat and blood lipids within individuals. Finally, in model
10, the 14 cross-trait correlations of the interindividual matrices
(see Table 2) are simultaneously fixed at zero to test the
hypothesis that there is no cross-trait resemblance in the
interindividual correlations. The most parsimonious model was obtained
by combining all nonrejected hypotheses in a single test.
All these hypotheses were tested using the likelihood ratio test, which
is minus twice the difference in the log-likelihoods obtained under two
different models. The ratio is distributed approximately as a
2 with the degrees of freedom being the
difference in the number of parameters estimated in the two
competing (nested) hypotheses.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Table 3
summarizes the pattern of
significant intra- and interindividual cross-trait correlations found
between the four body fat and five lipid phenotypes. In this
Table, "Yes" designates that the spouse, parent-offspring, and
sibling cross-trait correlations are significant (P<.05);
"No" indicates that the correlations are nonsignificant; and
"Some" indicates that at least one of these correlations is
significant.
|
The first general pattern noted is the presence of significant
intraindividual and interindividual cross-trait resemblance between the
two body fat measures (BMI and SF6) and lipid variables. The only
exception to this pattern is the absence of significant interindividual
cross-trait resemblance with CH. On the other hand, there are few
significant cross-trait correlations between fat distribution measures
(TER and TER-sf) and the lipids, especially after adjustment for amount
of subcutaneous fat (TER-sf). In the latter case, cross-trait
resemblance is found only with CH/HDL and CH-HDL. For these pairs of
measures, a general pattern of significant (or borderline significant)
sibling or sibling and parent-offspring cross-trait correlations are
noted (see footnotes in Table 3 for details).
Table 4 presents the maximum
likelihood estimates of the intraindividual cross-trait correlations
derived from the most parsimonious model. Intraindividual correlations
involving BMI and SF6 were all significant, suggesting, as expected,
that body fat and blood lipids are correlated within individuals. The
general pattern of correlations suggests that blood lipids are more
strongly associated with indicators of body fat (positive correlations
with CH, TG, CH/HDL, and CH-HDL and negative correlations with HDL)
than indicators of body fat distribution. Indeed, when TER is adjusted
for amount of subcutaneous fat (TER-sf), the correlations are
essentially nonsignificant and equal to zero.
|
Table 5 presents the interindividual
sibling cross-trait correlations estimated under the most parsimonious
model. For all body fat-lipid pairs of measures, the spouse and
parent-offspring cross-trait correlations were not significant and
hence fixed at zero in the parsimonious model. SF6 is the body fat
indicator showing the most consistent cross-trait resemblance
with blood lipids. Indeed, except for CH, all the blood lipid
variables exhibited cross-trait sibling resemblance with SF6. The
presence of cross-trait sibling resemblance in the absence of
cross-trait parent-offspring resemblance suggests that genetic factors
are probably not as important as environmental factors in determining
the covariation between body fat and blood lipids. However, in the
absence of significant cross-trait spouse correlations, another
explanation could be that pleiotropic effects of genes affecting body
fat and blood lipids are transient and less important at older
ages.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The search for familial resemblance in the covariation between pairs of phenotypes involved in the metabolic syndrome represents a quick and efficient way to investigate the shared genetic basis of the various metabolic disturbances characterizing this syndrome. Computation of bivariate familial correlations requires few assumptions and examination of the pattern of correlations among spouses, parent-offspring, and siblings could lead to the same genetic and environmental inferences as those derived from univariate familial correlations. For example, significant parent-offspring and sibling cross-trait correlations with no spouse correlation would suggest genetic pleiotropy, ie, that body fat and blood lipids may be influenced by common genes. Significant spouse correlations in addition to parent-offspring and sibling correlations suggest that familial environment could also contribute to the covariation observed between traits.
Based on likelihood ratio tests, evidence of significant cross-trait resemblance was found between indicators of body fat and fat distribution and the blood lipid variables. Examination of these cross-trait correlations reveals a more consistent pattern of cross-trait familial resemblance between body fat (BMI and SF6), rather than fat distribution (TER-sf), and blood lipids. Except for CH, all lipid measures were found to share genetic and/or environmental factors with body fat, while only CH/HDL and CH-HDL showed cross-trait resemblance with fat distribution. These two lipid ratios, used as indices of non-HDL cholesterol and apoB-associated cholesterol, respectively, are more closely associated with atherosclerosis and the risk of coronary heart disease than total cholesterol. Recent studies, for example, have shown that elevated apoB levels were associated with a threefold increase in the risk of ischemic heart disease28 and that fasting hyperinsulinemia combined with elevated apoB levels was associated with more than a tenfold increase in the risk of ischemic heart disease.29 These results suggest that the clustering of lipoprotein abnormalities commonly associated with obesity, especially upper-body obesity, is clearly atherogenic. The finding of pleiotropic effects between TER-f and CH-HDL in the present study suggest that genetic and/or environmental factors could contribute to this clustering.
The finding of significant intraindividual but not interindividual
cross-trait correlations emphasizes the need to distinguish between
these two types of correlations in genetic analyses. The
presence of an intraindividual correlation between two variables
like body fat and blood lipids, even when each of them are
significantly influenced by genetic factors, does not imply a shared
genetic basis. Rather, the observed cross-trait correlations within
individuals may be the result of specific environmental factors which
are unique to each individual and, thus, not necessarily heritable.
This is the situation likely prevailing in the present study since,
for most of the body fat-blood lipid pairs of measures, the
interindividual cross-trait correlations are low. A closer look at the
pattern of intraindividual cross-trait resemblance observed in this
study (see Table 4) reveals that not all measures of body fat are
equally related to blood lipids and lipoproteins within individuals.
For example, the amount of subcutaneous fat assessed by the sum of 6
skin folds (SF6) was found to be correlated with all lipid
phenotypes, but no significant correlations were observed with
the proportion of subcutaneous body fat found on the trunk after
adjustment for the amount of subcutaneous fat (see TER-sf in Table 4).
This finding suggests that the amount of subcutaneous fat, which is
strongly correlated with total body fat, is a better correlate of the
lipid phenotypes than indicators of subcutaneous fat
distribution.
The significant interindividual cross-trait resemblance observed
between body fat and blood lipids is almost exclusively found in the
sibling and not in the parent-offspring correlations, which is not
compatible with a simple genetic effect. One explanation for this
pattern of cross-trait correlations may be that some environmental
factors shared between siblings but not between parents and their
offsprings contribute to the cross-trait familial resemblance. Factors
such as activity and nutritional habits may be involved. Despite this
pattern of cross-trait resemblance, the hypothesis of a shared genetic
basis between body fat and blood lipids cannot be completely ruled out,
since significant cross-trait sibling but not parent-offspring
correlations would be expected under the hypothesis that genes
influencing the covariation between body fat and blood lipids are
age-dependent and transient. In the present study, the siblings
were young (15 years old on average), which would support the
hypothesis of a transient genetic effect in the covariation between
body fat and blood lipids. Assuming that both common genes and
environmental factors contribute to this cross-trait resemblance,
bivariate familiality may be approximated by simply doubling the
average cross-trait sibling correlations. Based on the cross-trait
correlations presented in Table 5, the bivariate familiality
would reach 12% between BMI and TG and 8% between BMI and both CH/HDL
and CH-HDL. For SF6-TG, which evidenced significant sex differences,
the bivariate familiality may be as high as 40% in females and 14% in
males, while, for SF6-CH/HDL, the cross-trait heritability would be
8%.
Relatively few studies have looked for pleiotropic effects between adiposity measures and blood lipids and lipoproteins. In one study conducted on 665 individuals from 135 kindreds, Towne et al,30 estimated the additive genetic correlation between CH and BMI as well as WHR and found that this correlation was not significantly different from zero between either BMI and CH or between WHR and CH. In another study based on data from 2184 households comprising 5376 individuals living in Gubbio, Italy, Schork et al,31 reported pleiotropic effects between BMI and levels of CH and HDL, but with significant contribution of pleiotropic genes only for the covariation between BMI and CH. Recently, two multivariate genetic analyses studies based on data from the San Antonio Family Heart Study investigated the contribution of shared genetic and environmental factors among traits related to the metabolic syndrome.32,33 In one of these studies, common environmental factors, rather than shared genes, were responsible for the covariation of plasma levels of TG and HDL with BMI or fat mass estimated by bioelectrical impedance.32 These results are in agreement with those reported in the present study and suggest that shared environmental factors are contributing more strongly to the covariation observed between body fat and blood lipids than shared genetic factors.
In summary, the results of this study indicate the presence of significant cross-trait resemblance between body fat and blood lipids. The cross-trait interindividual resemblance was found to be almost exclusively accounted for by significant sibling correlations rather than parent-offspring correlations. Although a shared genetic basis between body fat and blood lipid variation cannot be definitely ruled out from the pattern of cross-trait familial resemblance, the results suggest that environmental factors specific to each individual and common familial environmental factors are probably more important than genetic factors in explaining the covariation observed between body mass, body fat, fat distribution, and blood lipids. In the aggregate, the results of the present study support those of previous multivariate genetic studies of the metabolic syndrome and suggest that both shared genetic and environmental factors contribute to the clustering of the risk factors which characterize the metabolic syndrome.
|
Selected Abbreviations and Acronyms |
|---|
|
|
|
Acknowledgments |
|---|
This work was partly supported by NIH grants GM-28719, and MRC of Canada Grant PG-11811. Thanks are expressed to the members of the Physical Activity Sciences Laboratory and Lipid Research Center who were involved in the data collection for this study. Finally, we are grateful to all families who gave their time to participate to the Quebec Family Study.
Received December 20, 1996; accepted June 16, 1997.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Donahue RP. The insulin resistance syndrome (syndrome X) and risk factors for coronary heart disease: an epidemiologic overview. Endocrinologist. 1994;4:112–116.
2. Reaven GM. Role of insulin resistance in human disease. Diabetes. 1988;37:1595–1607.[Abstract]
3.
Kaplan NM. The deadly quartet. Upper-body obesity,
glucose intolerance, hypertriglyceridemia
and hypertension. Arch Intern Med. 1989;149:1514–1520.
4. DeFronzo RA, Ferrannini E. Insulin resistance. A multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease. Diabetes Care. 1991;14:173–194.[Abstract]
5. Björntorp P. Visceral obesity: a `civilization syndrome.' Obesity Res. 1993;1:206–222.[Medline] [Order article via Infotrieve]
6. Crepaldi G, Tiengo A, Manzato E, eds. The plurimetabolic syndrome. Diabetes, obesity and hyperlipidemias. Amsterdam, Netherlands: Excerpta Medica, 1993;5.
7. Ducimetiere P, Richard J, Cambien F. The pattern of subcutaneous fat distribution in middle-aged men and the risk of coronary heart disease: the Paris prospective study. Int J Obes. 1986;10:229–240.[Medline] [Order article via Infotrieve]
8. Zamboni M, Armellini F, Sheiban I, Marchi MD, Todesco T, BeAndreis IA, Cominacini L, Bosello O. Relation of body fat distribution in men and degree of coronary narrowings in coronary artery disease. Am J Cardiol. 1992;70:1135–1138.[Medline] [Order article via Infotrieve]
9. Hauner H, Stangl K, Schmatz C, Burger K, Blomer H, Pfeiffer EF. Body fat distribution in men with angiographically confirmed coronary artery disease. Atherosclerosis. 1990;85:203–210.[Medline] [Order article via Infotrieve]
10. Hauner H, Bognar E, Blum A. Body fat distribution and its association with metabolic and hormonal risk factors in women with angiographically assessed coronary artery disease: evidence for the presence of a metabolic syndrome. Atherosclerosis. 1994;105:209–216.[Medline] [Order article via Infotrieve]
11. Hodgson JM, Wahlqvist ML, Balazs NDH, Boxall JA. Coronary atherosclerosis in relation to body fatness and its distribution. Int J Obes. 1994;18:41–46.
12. Kissebah AH, Peiris AN. Biology of regional body fat distribution: relationship to non-insulin-dependent diabetes mellitus. Diabetes Metab Rev. 1989;5:83–109.[Medline] [Order article via Infotrieve]
13.
Després JP, Moorjani S, Lupien PJ, Tremblay A,
Nadeau A, Bouchard C. Regional distribution of body fat, plasma
lipoproteins, and cardiovascular disease.
Arteriosclerosis. 1990;10:497–511.
14. Després JP. Obesity and lipid metabolism: relevance of body fat distribution. Curr Opin Lipidol. 1991;2:5–15.
15. Bouchard C, Pérusse L. Genetics of causes and manifestations of the metabolic syndrome. In: Crepaldi G, Tiengo A, Mamzato E, eds. Diabetes, obesity and hyperlipidemia, V: The plurimetabolic syndrome. Amsterdam, Netherlands: Elsevier Science Pub; 1993:67–74.
16. Bouchard C, Pérusse L, Leblanc C, Tremblay A, Thériault G. Inheritance of the amount and distribution of human body fat. Int J Obes. 1988;12:205–215.[Medline] [Order article via Infotrieve]
17.
Pérusse L, Després J P, Tremblay A, Leblanc
C, Talbot C, Allard C, Bouchard C. Genetic and environmental
determinants of serum lipids and lipoproteins in French Canadian
families. Arteriosclerosis. 1989;9:308–318.
18. Carmelli D, Cardon LR, Fabsitz R. Clustering of hypertension, diabetes, and obesity in adult male twins: same genes or same environments? Am J Hum Genet. 1994;55:566–573.[Medline] [Order article via Infotrieve]
19. Hong Y, Pedersen NL, Brismar K, de Faire U. Genetic and environmental architecture of the features of the insulin-resistance syndrome. Am J Hum Genet. 1997;60:143–152.[Medline] [Order article via Infotrieve]
20. Rice T, Province M, Pérusse L, Bouchard C, Rao DC. Cross-trait familial resemblance for body fat and blood pressure: familial correlations in the Québec family study. Am J Hum Genet. 1994;55:1019–1029.[Medline] [Order article via Infotrieve]
21. Rice T, Nadeau A, Pérusse L, Bouchard C, Rao DC. Familial correlations in the Quebec family Study: cross-trait familial resemblance for body fat with plasma glucose and insulin. Diabetologia. 1996;39:1357–1364.[Medline] [Order article via Infotrieve]
22. Bouchard C. Genetic epidemiology, association, and sib-pair linkage: results from the Quebec family study. In: Bray GA, Ryan DH, eds. Molecular and genetic aspects of obesity. Pennington Center Nutrition Series. Baton Rouge, La: Louisiana State University Press; 1996;5:470–481.
23. Weiner JS, Lourie JA. Human biology: a guide to field methods. Oxford, London: Blackwell Scientific Publications, 1969.
24. Leclerc S, Bouchard C, Talbot J, Gauvin R, Allard C. Association between serum high-density lipoprotein cholesterol and body composition in adult men. Int J Obes. 1983;7:555–561.[Medline] [Order article via Infotrieve]
25. Borecki IB, Rice T, Bouchard C, Rao DC. Commingling analysis of generalized body mass and composition measures: the Quebec Family Study. Int J Obes. 1991;15:763–773.[Medline] [Order article via Infotrieve]
26. Rice T, Borecki IB, Bouchard C, Rao DC. Commingling analysis of regional fat distribution measures: the Quebec family study. Int J Obes. 1992;16:831–844.
27. Province MA, Rao DC. General purpose model and a computer program for combined segregation and path analysis (SEGPATH): automatically creating computer programs from symbolic language model specifications. Genet Epidemiol. 1995;12:203–219.[Medline] [Order article via Infotrieve]
28.
Lamarche B, Moorjani S, Lupien PJ, Cantin B, Dagenais
GR, Després JP. Apolipoprotein A-I and B levels and the risk of
ischemic heart disease during a 5-year follow-up of men in the
Quebec cardiovascular study. Circulation. 1996;94:273–278.
29.
Després JP, Lamarche B, Mauriège P,
Cantin B, Dagenais GR, Moorjani S, Lupien PJ.
Hyperinsulinemia as an independent risk factor for
ischemic heart disease. N Engl J Med. 1996;334:952–957.
30. Towne B, Roche AF, Chumlea WC, Guo S, Siervogel RM. No evidence of pleiotropy for either body mass index or waist/hip circumference ratio and plasma cholesterol concentration. Obesity Res. 1993;1:1105.
31. Schork NJ, Weder AB, Trevisan M, Laurenzi M. The contribution of pleiotropy to blood pressure and body-mass index variation: the Gubbio study. Am J Hum Genet. 1994;54:361–373.[Medline] [Order article via Infotrieve]
32.
Mahaney MC, Blangero J, Comuzzie G, VandeBerg JL, Stern
MP, MacCluer JW. Plasma HDL cholesterol,
triglycerides, and adiposity. A quantitative genetic test
of the conjoint trait hypothesis in the San Antonio family heart study.
Circulation. 1995;92:3240–3248.
33.
Mitchell BD, Kammerer CM, Mahaney C, Blangero J,
Comuzzie AG, Atwood LD, Haffner SM, Stern MP, MacCluer JW. Genetic
analysis of the IRS. Pleiotropic effects of genes influencing
insulin levels on lipoprotein and obesity measures. Arterioscler
Thromb Vasc Biol. 1996;16:281–288.
This article has been cited by other articles:
 |
|
 |
|
  S. B. Seidelmann, C. De Luca, R. L. Leibel, J. L. Breslow, A. R. Tall, and C. L. Welch Quantitative Trait Locus Mapping of Genetic Modifiers of Metabolic Syndrome and Atherosclerosis in Low-Density Lipoprotein Receptor-Deficient Mice: Identification of a Locus for Metabolic Syndrome and Increased Atherosclerosis on Chromosome 4 Arterioscler Thromb Vasc Biol, January 1, 2005; 25(1): 204 - 210. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C.Y. Ng, W.-Y. So, V. K.L. Lam, C. S. Cockram, G. I. Bell, N. J. Cox, and J. C.N. Chan Genome-wide Scan for Metabolic Syndrome and Related Quantitative Traits in Hong Kong Chinese and Confirmation of a Susceptibility Locus on Chromosome 1q21-q25 Diabetes, October 1, 2004; 53(10): 2676 - 2683. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. F. Loos, P. T. Katzmarzyk, D. C. Rao, T. Rice, A. S. Leon, J. S. Skinner, J. H. Wilmore, T. Rankinen, and C. Bouchard Genome-Wide Linkage Scan for the Metabolic Syndrome in the HERITAGE Family Study J. Clin. Endocrinol. Metab., December 1, 2003; 88(12): 5935 - 5943. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. T. Katzmarzyk, N. Gledhill, L. Perusse, and C. Bouchard Familial Aggregation of 7-Year Changes in Musculoskeletal Fitness J. Gerontol. A Biol. Sci. Med. Sci., December 1, 2001; 56(12): B497 - 502. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B. Harrap, M. Stebbing, J. L. Hopper, H. N. Hoang, and G. G. Giles Familial Patterns of Covariation for Cardiovascular Risk Factors in Adults: The Victorian Family Heart Study Am. J. Epidemiol., October 15, 2000; 152(8): 704 - 715. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||